NORMA eResearch @NCI Library

Optimized Pre-Copy Live Virtual Machine Migration for Memory-Intensive Workloads

Jain, Prateek (2021) Optimized Pre-Copy Live Virtual Machine Migration for Memory-Intensive Workloads. Masters thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Master of Science)
Download (1MB) | Preview
[img]
Preview
PDF (Configuration manual)
Download (2MB) | Preview

Abstract

With the expansion of IT infrastructures and rising energy and power costs, maintaining workload concentration and the high availability of virtual machines (VMs) in a data center is becoming increasingly difficult. To mention a few issues, some physical servers may be overloaded, while others may be idle. If a server fails, all VMs on it are affected; and so on. To guarantee service continuity, these impacted VMs must be moved to other servers. These issues (how to equally allocate jobs among servers, how to safeguard VMs from equipment failures, and so on) are being addressed in tandem with the introduction of a crucial technology—VM migration. Pre-copy migration is a popular method for transferring VMs across physical servers. Before terminating the VM on the source, pre-copy moves the VM’s memory state from source to destination in successive iterations. Although the pre-copy approach reduces downtime and total migration time, it does limit the number of copying rounds. Because the writable working set is not guaranteed to converge across consecutive cycles, especially when the VM is running a predominantly write-intensive application, an alternative and more precise method are necessary to manage memory-intensive workloads. The improved migration method is provided in this work to address the limitations of the pre-copy migration approach. The technique works in conjunction with KVM’s default migration mechanism. Memcached, a key-value store application is used to assess the performance of the improved VM migration mechanism. Oracle VirtualBox is used as a test environment to carry out the migration procedure.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science

T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions: School of Computing > Master of Science in Cloud Computing
Depositing User: Clara Chan
Date Deposited: 13 Oct 2021 17:45
Last Modified: 13 Oct 2021 17:52
URI: http://norma.ncirl.ie/id/eprint/5087

Actions (login required)

View Item View Item